Presentation is loading. Please wait.

Presentation is loading. Please wait.

Prediction of in-hospital mortality after ruptured abdominal aortic aneurysm repair using an artificial neural network  Eric S. Wise, MD, Kyle M. Hocking,

Similar presentations


Presentation on theme: "Prediction of in-hospital mortality after ruptured abdominal aortic aneurysm repair using an artificial neural network  Eric S. Wise, MD, Kyle M. Hocking,"— Presentation transcript:

1 Prediction of in-hospital mortality after ruptured abdominal aortic aneurysm repair using an artificial neural network  Eric S. Wise, MD, Kyle M. Hocking, PhD, Colleen M. Brophy, MD  Journal of Vascular Surgery  Volume 62, Issue 1, Pages 8-15 (July 2015) DOI: /j.jvs Copyright © 2015 Society for Vascular Surgery Terms and Conditions

2 Fig 1 Schematic representation of the four-variable artificial neural network (ANN-4). The four variables on the left compose the input layer. The four processing elements (nodes) represent the ANN's hidden layer; a training value of was used. The output layer is a predicted likelihood of death, transformed from 0 (survival to discharge) to 1 (in-hospital mortality). Journal of Vascular Surgery  , 8-15DOI: ( /j.jvs ) Copyright © 2015 Society for Vascular Surgery Terms and Conditions

3 Fig 2 Four-variable multiple logistic regression model. A, Receiver operating characteristic (ROC) curve for the multiple logistic regression analysis with the four significant variables considered in the algorithm (n = 107 patients). The area under the curve (AUC) was 0.85 ± B, Actual vs predicted outcomes plot; the correlation between actual and predicted death by simple linear regression was r2 = .36. Root-mean-square error = The dotted lines represent the 95% confidence band. Axes: 0, survival (predicted or actual); 1, in-hospital mortality (predicted or actual). Journal of Vascular Surgery  , 8-15DOI: ( /j.jvs ) Copyright © 2015 Society for Vascular Surgery Terms and Conditions

4 Fig 3 Four-variable artificial neural network (ANN-4) model. A, Receiver operating characteristic (ROC) curve for the simplified ANN-4 training cohort (solid line, n = 86 patients; area under the curve [AUC] was 0.88 ± 0.04) and ANN-4 validation cohort (dashed line, n = 21 patients; AUC = 0.95 ± 0.06). B, Actual vs predicted outcomes plot; the correlation between actual and predicted death by simple linear regression on all 107 patients, using the training set-derived algorithm, was r2 = .52. Root-mean-square error = The dotted lines represent the 95% confidence band. Axes: 0, survival (predicted or actual); 1, in-hospital mortality (predicted or actual). Journal of Vascular Surgery  , 8-15DOI: ( /j.jvs ) Copyright © 2015 Society for Vascular Surgery Terms and Conditions

5 Fig 4 Patient in-hospital mortality expressed by Glasgow Aneurysm Score (GAS) stratum (n = 83 patients). Journal of Vascular Surgery  , 8-15DOI: ( /j.jvs ) Copyright © 2015 Society for Vascular Surgery Terms and Conditions

6 Fig 5 In-hospital mortality was correlated to the Glasgow Aneurysm Score (GAS; n = 83 patients). A, The receiver operating characteristic (ROC) curve is shown; area under the curve (AUC) is 0.77 ± B, Actual vs predicted outcomes plot; the correlation between actual and predicted death by simple linear regression was r2 = .17. Root-mean-square error (GAS normalized to 0-1 scale) = The dotted lines represent the 95% confidence band. Axes: 0, survival (predicted or actual); 1, in-hospital mortality (predicted or actual). Journal of Vascular Surgery  , 8-15DOI: ( /j.jvs ) Copyright © 2015 Society for Vascular Surgery Terms and Conditions


Download ppt "Prediction of in-hospital mortality after ruptured abdominal aortic aneurysm repair using an artificial neural network  Eric S. Wise, MD, Kyle M. Hocking,"

Similar presentations


Ads by Google